Ubiquitin ligase RNF20 coordinates sequential adipose thermogenesis with brown and beige fat-specific substrates

Abstract

In mammals, brown adipose tissue (BAT) and inguinal white adipose tissue (iWAT) execute sequential thermogenesis to maintain body temperature during cold stimuli. BAT rapidly generates heat through brown adipocyte activation, and further iWAT gradually stimulates beige fat cell differentiation upon prolonged cold challenges. However, fat depot-specific regulatory mechanisms for thermogenic activation of two fat depots are poorly understood. Here, we demonstrate that E3 ubiquitin ligase RNF20 orchestrates adipose thermogenesis with BAT- and iWAT-specific substrates. Upon cold stimuli, BAT RNF20 is rapidly downregulated, resulting in GABPα protein elevation by controlling protein stability, which stimulates thermogenic gene expression. Accordingly, BAT-specific Rnf20 suppression potentiates BAT thermogenic activity via GABPα upregulation. Moreover, upon prolonged cold stimuli, iWAT RNF20 is gradually upregulated to promote de novo beige adipogenesis. Mechanistically, iWAT RNF20 mediates NCoR1 protein degradation, rather than GABPα, to activate PPARγ. Together, current findings propose fat depot-specific regulatory mechanisms for temporal activation of adipose thermogenesis.

Fig. 1. Upon Acute Cold, BAT RNF20 Downregulation Potentiates Thermogenic Activity.

a Western blot analysis of a differentiated brown preadipocyte cell line (BAC) treated with isoproterenol (Iso. 1 μM) and relative band intensity normalised to ACTIN. The p-value was calculated compared to the 0 h group. n = 6 independent replicates. Representative results from two independent experiments. b Western blot analysis of brown adipose tissue (BAT) from mice housed at room temperature (RT, n = 4 mice) and exposed to a cold environment (6 °C, n = 5 (6 h), n = 4 mice (72 h)). Representative results from three independent experiments. The p-value was calculated compared to the 0 h group. c, d qRT-PCR and western blotting analyses of differentiated BAC cells transfected with mock or RNF20-expressing plasmid. n = 3 independent replicates. Ct: critical threshold. e Oxygen consumption rate (OCR) of differentiated brown adipocytes transfected with mock or RNF20-expressing plasmid. n = 5 (Mock), n = 4 mice (RNF20 OE). f, g qRT-PCR and western blot analyses of differentiated BAC cells transfected with siNC or siRnf20. n = 3 independent replicates. h OCR of differentiated BAC cells transfected with siNC or siRnf20. n = 5 independent replicates (siNC), n = 4 independent replicates (siRnf20). i Representative macroscopic view, wholemount MitoTracker staining images, and quantitative analysis of MitoTracker intensity of BAT of WT and Rnf20^+/− mice housed under RT. j–l Western blot analysis, qRT-PCR analysis (n = 4 (WT), n = 4 (Rnf20^+/−) mice), and mitochondrial DNA (mtDNA) (n = 7 (WT), n = 5 (Rnf20^+/−) mice) content of genes related to thermogenesis in BAT of WT and Rnf20^+/− mice housed under RT. Representative results from two independent experiments. m–p Rectal temperature and representative infrared images of male (n = 4 (WT), n = 4 (Rnf20^+/−) mice) and female (n = 8 (WT), n = 4 (Rnf20^+/−) mice) WT and Rnf20^+/− mice upon cold exposure (6 °C). For (m, n) representative results from three independent experiments. q, r Wholebody heat generation, oxygen consumption, physical activity, drink, and food intake in male WT and Rnf20^+/− mice housed under RT. During the experiment, β3-adrenergic agonist CL316,243 (CL, 0.5 mg/kg body weight) was administrated (arrow). n = 5 mice. Source data are provided as a Source Data file. n indicates biological replicates. Data are represented as mean ± S.D. Significance was determined using one-way ANOVA with Dunnett’s multiple comparison (a, b, q), unpaired two-sided Student’s t-test (d, e, g, h, i, j, k), repeated measures ANOVA with Tukey’s multiple comparisons test (m, o, q), and two-way ANOVA (q, r).

Fig. 2. BAT-specific RNF20 Modulation Controls Thermogenic Activity Upon Acute Cold Stimuli.

a Experimental scheme for plasmid injection (5 μg) into BAT with in vivo jetPEI. b qRT-PCR of BAT and other peripheral tissues in mock and RNF20 overexpression (OE) mice. n = 5 (Mock), n = 3 (RNF20 OE) mice. c–f Western blotting, qRT-PCR, mtDNA content of BAT from mock and RNF20 OE mice housed under RT. n = 4 (Mock), n = 6 (RNF20 OE) mice. Representative results from two independent experiments. g–j Rectal temperature and representative infrared images of mock and RNF20 OE male (n = 6 (Mock), n = 6 (RNF20 OE) mice) and female mice (n = 8 (Mock), n = 7 (RNF20 OE) mice) upon cold exposure (6 °C). Representative results from three independent experiments. k Representative H&E images of BAT from mock and RNF20 OE mice and the distribution of the LD size of BAT. l Representative image showing BAT of RNF20 OE mice floating in phosphate-buffered saline (PBS; density: 1.0723 g/cm³). m Experimental scheme for siRNA injection (5 μg) into BAT. n Western blot analysis of BAT of siNC or siRnf20 mice housed under RT. Representative results from two independent experiments. o–r Rectal temperature and representative infrared images of siNC or siRnf20 male (n = 5 (siNC), n = 4 (siRnf20) mice) and female mice (n = 7 (siNC), n = 8 (siRnf20) mice) upon cold exposure (6 °C). In vivo experiments were performed 3 d after nucleotide injection. Representative results from two independent experiments. Source data are provided as a Source Data file. n indicates biological replicates. Data are represented as mean ± S.D. Significance was determined using unpaired two-sided Student’s t-test (b, d, f), multiple unpaired t-test with False Discovery Rate (b), and repeated measures ANOVA with Tukey’s multiple comparisons test (g, i, o, q).

Fig. 3. In BAT, RNF20 Regulates Thermogenesis via GABPα.

a–c Volcano plot, gene ontology, and heatmap of the proteome from WT and Rnf20^+/− mice. n = 3 mice. d Crucial transcription factors (TFs) of upregulated proteins. e Gene network analysis of upregulated proteins in BAT of WT and Rnf20^+/− mice. GABPα target genes are labelled as red connection lines. f qRT-PCR analysis of GABPα-targeted mitochondrial genes in the BAT of WT and Rnf20^+/− mice housed at RT. n = 7 (WT), n = 5 (Rnf20^+/−) mice. g Gene clustering analysis of the BAT transcriptome (GSE119452). h Gene ontology and crucial transcription factors of the upregulated genes (cluster 2). i t-distributed stochastic neighbour embedding (tSNE) plot of BAT single-cell RNA-seq data (GSE125269). Adipocytes were largely divided into high^thermo BAs, low^thermo BAs, and white adipocytes (WAs). j Gene ontology and crucial transcription factors of high^thermo BA-enriched genes. k Mouse Ucp1 enhancer regions with H3K27Ac enrichment peaks in BAT (GSE63964). Primers used for enhancer cloning are indicated. l Luciferase activity of Ucp1-luciferase (luc.) constructs containing their enhancer (−2.5 kb upstream, Materials and Methods), minimal promoter (Mini prom), and TATA-box element. n = 3 independent replicates. m Correlation of GABPA expression level with UCP1-enhancer luciferase activity in differentiated clonal brown preadipocytes from human BAT (GSE68544). n, o Western blotting and rectal temperatures during cold exposure (6 °C) of mock and GABPA-OE mice housed under RT. n = 3 mice. Representative results from two independent experiments. p Western blotting analysis of BAT of WT, Rnf20^+/−, WT with siGabpa, and Rnf20^+/− with siGabpa mice housed under RT. q Rectal temperatures of WT (n = 5), Rnf20^+/− (n = 4), WT with siGabpa (n = 6), and Rnf20^+/− with siGabpa (n = 4) mice during cold exposure (6 °C). In vivo experiments were performed 3 d after nucleotide injection. Source data are provided as a Source Data file. n indicates biological replicates. Data are represented as mean ± S.D. Significance was determined using multiple unpaired t-tests with False Discovery Rate (a, f), Fisher’s exact test with Benjamini–Hochberg method (b, d, h, j), unpaired two-sided Student’s t-test (l), two-tailed Pearson correlation (m), repeated measures ANOVA with Tukey’s multiple comparisons test (o), and two-way ANOVA (p, q).

Fig. 4. RNF20 Promotes GABPα Protein Degradation in BAT, A Crucial Factor for Thermogenesis.

a Western blotting analysis of inguinal WAT (iWAT) and BAT. b, c Protein and mRNA levels (n = 4 (RT), n = 4 (Cold) mice) of Gabpa in BAT from mice upon cold exposure (6 °C) for 6 h. d, e Protein and mRNA levels (n = 4 (Mock), n = 6 (RNF20 OE) mice) of Gabpa in BAT of mock and RNF20 OE mice. f, g Protein and mRNA levels (n = 7 (WT), n = 5 (Rnf20^+/−) mice) of Gabpa in BAT of WT and Rnf20^+/− mice. h, i Protein and mRNA levels of Gabpa (n = 6 (siNC), n = 6 (siRnf20) mice) in BAT of mice injected with siNC or siRnf20. For a–i representative results from two independent experiments. j Endogenous co-immunoprecipitation using GABPα antibody in BAT. k Ubiquitin levels of GABPα in BAT of mock and RNF20 OE. l Ubiquitin levels of GABPα in BAT of siNC and siRnf20. Representative results from two independent experiments. m Protein level of GABPα transfected with RNF20-expressing plasmids in HEK293T cells without or with MG132 treatment (20 μM for 6 h). n, o Cycloheximide (30 μM)-chasing assay of GABPα protein in differentiated brown adipocytes. n = 3 independent replicates. Representative results from three independent experiments. o Scheme for luciferase assay using mouse Mtif2 promoter (−86 bp upstream) in which GABPα-binding motifs (C/AGGAAG or CTTCCT/G) are present (GABPα activity luc.). p GABPα activity luc. with GABPα and GABPβ1-expressing plasmids. n = 3 independent replicates. q–s Effects of RNF20 overexpression and knockdown on GABPα or Ucp1-luciferase activity. n = 3. Source data are provided as a Source Data file. n indicates biological replicates. Data are represented as mean ± S.D. Significance was determined using an unpaired two-sided Student’s t-test (c, e, g, i), and one-way ANOVA with Tukey’s multiple comparison (m–o, q–s).

Fig. 5. In iWAT, RNF20 Potentiates Beige Fat Thermogenesis upon Prolonged Cold Stimuli, Independent of GABPα.

a Representative H&E staining of inguinal white adipose tissue (iWAT) from mice exposed to a cold environment (6 °C) for 6, 24, and 72 h. Scale bar: 50 μm. Representative results from two independent experiments. b Protein levels of RNF20 and UCP1 in iWAT during cold exposure n = 5 (RT), n = 5 (6 h), n = 5 (72 h) mice. Representative results from two independent experiments. c Experimental scheme for iWAT-specific RNF20 overexpression (5 μg). d–h RNF20 protein level, representative infrared image, iWAT region temperatures (n = 5 for male and n = 4 for female mice), representative H&E staining, qRT-PCR (n = 6 mice), and western blotting analysis of iWAT from mock and RNF20 OE mice exposed to cold (6 °C) for 72 h. Representative results from two independent experiments. i Experimental scheme for iWAT-specific siRNA treatment (5 μg). j, k iWAT region temperatures (n = 7 for male and n = 6 for female mice) and qRT-PCR analysis (n = 6 mice) of iWAT from siNC and siRnf20 mice exposed to cold (6 °C) for 72 h. l, m GABPα mRNA and protein levels in iWAT from iWAT RNF20 OE mice exposed to cold (6 °C) for 72 h. n Ubiquitin levels of GABPα in iWAT from mock and RNF20 OE mice. Representative results from two independent experiments. Source data are provided as a Source Data file. n indicates biological replicates. Data are represented as mean ± S.D. Significance was determined using one-way ANOVA with Dunnett’s multiple comparison (b) and paired two-sided Student’s t-test (e, g, j, k, l, m).

Fig. 6. iWAT RNF20 Potentiates Beige Fat Thermogenesis by Promoting NCoR1 Degradation to Stimulate PPARγ.

a Gene ontology and transcription factor analysis of upregulated genes in iWAT transcriptome and expression levels of PPARγ target genes in iWAT upon 3-day cold stimuli (6 °C, GSE179385). b Protein levels of NCoR1 in iWAT during cold exposure (6 °C). n = 5 (RT), n = 5 (6 h), n = 5 (72 h) mice. Representative results from two independent experiments. c–e Ncor1 mRNA expression level, representative infrared image, iWAT region temperatures (n = 4 mice), and western blot analyses of iWAT from mock and NCoR1 OE mice (5 μg) exposed to cold (6 °C) for 3 d. f, g Western blot analysis and qRT-PCR (n = 6 mice) of iWAT from mock and RNF20 OE mice exposed to cold (6 °C) for 3 days. h–j Ncor1 and Rnf20 mRNA expression levels, representative infrared image, iWAT region temperatures (n = 4 mice), and western blotting analyses of iWAT from mock and RNF20&NCoR1 OE mice (5 μg for each gene) exposed to cold (6 °C) for 3 d. Source data are provided as a Source Data file. n indicates biological replicates. Data are represented as the mean ± S.D. Significance was determined using Fisher’s exact test (a), unpaired two-sided Student’s t-test (a) and paired two-sided Student’s t-test (c–e, g–j).

Fig. 7. iWAT RNF20 potentiates de novo beige adipogenesis upon prolonged cold stimuli.

a Gene ontology of upregulated genes in iWAT transcriptome upon 3-day cold stimuli (6 °C, GSE179385). b Western blot analysis of differentiated beige adipocytes from iWAT PDGFRα-expressing preadipocytes n = 3 (0 h), n = 3 (24 h), n = 3 (48 h), n = 3 (72 h) independent replicates. Representative results from two independent experiments. c–h Experimental scheme, a representative image of BODIPY lipid staining and qRT-PCR analysis of differentiated beige adipocytes from iWAT PDGFRα-expressing preadipocytes transduced with mock or RNF20-expressing plasmids (c–e) and siNC or siRnf20 (f–h) before adipogenic induction. i, j Experimental scheme using AdipoChaser mice. After 2 weeks of a diet containing doxycycline for YFP labelling, a normal diet was fed. After injection of mock or RNF20 into the left or right iWAT flank (5 μg), respectively, AdipoChaser mice were exposed to cold (6 °C). k Representative whole-mount images of iWAT from AdipoChaser mice upon cold exposure (6 °C) for 3 d. The asterisk (*) refers to newly differentiated YFP-negative adipocytes. AD adipocytes. Source data are provided as a Source Data file. n indicates biological replicates. Data are represented as mean ± S.D. Significance was determined using Fisher’s exact test (a), one-way ANOVA with Dunnett’s multiple comparison (b) and unpaired two-sided Student’s t-test (d, e, g, h, k).

Fig. 8. Proposed model.

Upon cold stimuli, mammals maintain their body temperature by potentiating adipose thermogenesis in a fat-depot-specific manner. In BAT, acute cold stimuli rapidly downregulate RNF20, which induces GABPα accumulation and activation. In turn, GABPα enhances thermogenesis by stimulating the expression of thermogenic and mitochondrial genes. Given that RNF20-NCoR1 axis in BAT would primarily regulate lipid metabolism, NCoR1 in BAT is not marked. In iWAT, prolonged cold stimuli gradually upregulate RNF20 to stimulate PPARγ by degrading NCoR1, which promotes de novo beige adipogenesis.